System for implementing fine grit classification and management of Web services

Abstract

The invention provides a system for implementing fine grit classification and management of Web services. The system is arranged on the boundary of a network and adopts the following implementing manner that known Web services are respectively trained by historical behavioral data to construct a behavioral model; the behavioral model is utilized to carry out online fine grit classification and management on the Web services; and for unknown types of Web services, the system automatically constructs a behavioral model and implements dynamic maintenance. The system is used for carrying out fine grit classification on different Web services which use the same HTTP (Hyper Text Transfer Protocol) and carrying out effective scheduling and management on HTTP streaming generated by different Web services according to a given service strategy. According to the invention, on the basis of HTTP session properties, not only does application layer payload analysis not need to be carried out, but also the defects of a conventional method on the basis of streaming and ports can be overcome, and fine classification and management of the Web services can be effectively implemented. The system has capabilities of dynamically updating, creating and refining. The system can be applied to various scenes with different targets.

Description

technical field [0001] The invention belongs to the field of network technology, and provides a system for realizing fine-grained classification and management of Web services. Background technique [0002] The World Wide Web (WWW), referred to as the Web, is a system for accessing interlinked hypertext documents through the network. In Web-based applications, the client only needs a browser to complete the task of requesting and sending data. This cross-platform feature and thin client feature make Web technology quickly penetrate into every application field, and play an important role in the modern Internet, such as: e-commerce, e-government and financial services. Due to the widespread use of web services, conventional network security management systems generally open the standard TCP port 80 used by web services, and by default the network traffic communicating with TCP port 80 is legal web services, that is, web browsing services. This feature has attracted more and...

AI Technical Summary

Problems solved by technology

But its main problems include: (1) It is not suitable for real-time online operation

Because this method needs to analyze and match the application layer payload, the computational complexity is very high, and its efficiency is difficult to adapt to the high-speed network environment

(2) It is difficult to deal with encrypted protocols or private protocols

However, it has the following disadvantages: (1) It is difficult to realize the active discovery of new applications

For unknown new business, this method is difficult to deal with effectively

(2) The current technology generally only distinguishes "big categories", such as: HTTP, P2P, SMTP, etc., without considering fine-grained network business division, such as: distinguishing various network applications that exist under the HTTP protocol

(3) It is difficult to reflect the real Web business characteristics with "flow" as the classification unit

[0010] From the above analysis, it can be seen that the current main methods have the following problems in dealing with the classification of Web services based on the HTTP protocol: (1) The fine-grained classification of Web services is not considered; (2) The application layer of Web services is not used Behavioral characteristics; (3) Difficult to handle unknown web services

However, the cookie-based method also has some defects: (1) In order to ensure security, cookies are usually encrypted, so except for the server side, other monitoring points cannot effectively use cookies even if they intercept them

(2) A large number of existing studies have shown that cookies have many security risks, such as: leaking user privacy, being used by malicious attackers to threaten the interests of customers, etc.

However, there are still limitations in realizing the fine-grained classification of web services: they are designed for the characteristics of web browsing, and are not suitable for other non-web browsing web services

The diversity of modern HTTP streaming services and the insufficiency of existing technologies have brought new challenges to traditional network resource management planning, and also threatened the traffic management strategy at the network border.

Images